- When I was a young I worked for a firm in Hyderabad that manufactured and distributed explosives. A major concern of our firm was safety; if something could go wrong it would go wrong. That led to keeping each unit-operation small even though there are obvious scale economies in the explosives business. My experience in India, and witnessing several tragic incidents, reinforced the soundness of these principles. In what follows, John Boyd and I argue that the same principles hold true for banks. The social costs resulting from one incident at a large unit, or a group of connected units, can be catastrophic. R. Jagannathan
All firms are bound to make mistakes and those mistakes will be bigger the bigger the firms involved. When big firms also happen to be financial institutions, especially banks, everyone in the economy will pay for their errors. That is because financial institutions are intricately connected with every aspect of the economy and they can and will bring an economy to a grinding halt if they cease to function. This is what economists call a “negative externality” and the Great Depression was a great example of what that means in practical terms. It is why banks have always been more heavily regulated than other firms, and why they should be treated the same today. Big banks present special problems and require special handling.
Over the last decade, banking regulation has become, inadvertently, largely ineffective. To some extent, this has not been the fault of the Fed, the Comptroller of the Currency, or the other bank regulators. They were dealt a tough hand. Passage of the Graham-Leach-Bliley Act in 1999, allowed banks to become so big and complicated they became very difficult to manage, let alone regulate. Overall, the bank regulators did an acceptable if not brilliant job given the circumstances.
What really caused the breakdown of bank regulation was the incredible rise of “quasi-banks,” firms that don’t have bank in their title but collectivelyprovide banking services, compete with regular banks, and have taken over a large share of the market. In particular, large non-bank financial institutions – mortgage originators, insurance firms, investment banks, rating agencies etc. – collectively and massively invaded the turf of traditional banks. They invested in illiquid real estate through mortgage loans and transformed those loans through financial engineering into highly liquid and apparently safe securities, competing with banks for funding. The quasi-banks were not banks traditionally defined, and were not subject to banking regulation. However, they performed banking activities, took huge risks, and caused huge problems.
The problems originated with there being too much money in the system, resulting in extremely low interest rates. Investors began searching for “riskless” investment opportunities with better yields; and Wall Street was willing to provide these. It created new apparently-riskless securities with attractive returns, by transforming mortgage loans through financial engineering. Since meeting the demand for the apparently-riskless securities involved generating mortgage loans, there was downward pressure on underwriting standards for mortgage loans so as to maintain supply.
Investors probably believed that these firms were too big to fail (hereafter TBTF) in two senses. The first sense was wrong: although the institutions were large, and their failures entailed massive losses, they took massive bets and in fact lost big enough to fail. The second sense was that they were too large for the government to allow their failure and this proved correct.
Because of the massive consolidation in the financial services industry, we now have a number of financial intermediaries that have become very large and important relative to the size of the economy. These institutions are TBTF in this second sense: if they get in trouble, they will be bailed out by government. For the government to do otherwise is to run the risk of causing domino effects, contagion, and a general meltdown of the financial system. Based on historical experience, this outcome is unavoidable. In the many banking crises around the world over last three decades (there have been about 100) TBTF bailouts and attendant social costs have been present in almost every instance.
Like many other economists we believe that TBTF policies actually generate a sort of pernicious feedback loop, indirectly causing the problems they are designed to contain. This is the so called “moral hazard” problem. Knowing they will be protected by government if things go badly, large banks are willing to take more risk than they otherwise would have done. To make matters worse, the special protection afforded by TBTF status is highly valuable. Recent research suggests that banking firms will pay huge acquisition premia just to get into the TBTF size range.
But note. Even if there were no feedback loop and no “moral hazard” problem, very large banking firms would still present a serious dilemma for policy-makers. Even the best of managers sometimes make fatal errors. When that occurs in a TBTF bank it is costly not just to them but costly to society.
The too big to fail umbrella is actually much bigger than many realize. In 2006 it covered about 65% of intermediated financial assets in the United States, defining TBTF as the assets and guarantees of the 25 largest bank holding companies plus Fannie and Freddie. Our computation does not include the big investment banks and recent events have demonstrated that many of these firms are TBTF. If we were to include them, the percentage would be greater than 65%.
Our Proposed Cures
Our proposed cures are two in number. The first is to identify all firms that perform critical banking functions and regulate them as banks. The second is to set a size constraint that prohibits financial firms from getting too big.
The quasi-banks need to be identified and made fully subject to banking regulation, whatever their name or kind of charter. If it swims, flies and quacks, it should be deemed “duck” and dealt with as such. No exceptions. This will be difficult to implement in practice. First, technology has blurred traditional business lines; for examples, eBay now provides payments services like a bank, and General Electric has become a huge lending institution. Second, financial firms are deft at end-running regulation and have a long history of doing that. However, we cannot shrink from identification merely because the task is difficult. This regulation should be done by a single regulator to avoid games of playing off one regulator against another. (Others can debate whether that one regulator should be the Fed).
To begin the identification task we propose two organizing principles. First, any firm would be defined “a bank” if its liabilities are transaction-able; that is, if they provide payments services like checking accounts do. This principle would identify as “a bank” all firms that are currently chartered as commercial banks, savings and loan associations, mutual savings banks, credit unions and industrial credit banks. It would also identify money market funds with checking account privileges. We have no problem with their inclusion since the money market mutual funds recently require government salvation.
The second principle holds that any firm would be defined as “a bank” if it predominantly held (or guaranteed) risky financial assets, and funded those holdings with liquid debt liabilities. This principle would have identified as “a bank” AIG, Fannie Mae, Freddie Mac most state chartered loan companies, most investment banks, and possibly General Electric Credit and the captive auto finance companies. It would exclude securities exchanges and brokers because they are primarily pass-throughs and do not finance large holdings with short term debt. Similarly, stock mutual funds would be excluded. Life and casualty insurance firms would generally be considered banks, but those firms are heterogeneous and their actual operations would have to be examined. Mortgage bankers would be excluded so long as they operate as originators and did not take significant positions in mortgages.
Once banks are broadly and realistically redefined, we propose that they be classified into two groups: those that are TBTF and those that are not.
Our proposal, notable for its simplicity, would be a simple rule that defined TBTF by size. For example, “If you are a bank and your assets, plus the assets you guarantee, exceed a threshold (say, $100 billion) you are TBTF.” Such a rule would be arbitrary, and might miss some smaller but still critical TBTF banks. If the size criterion were adopted, however, it would allow of a simple, direct solution to the TBTF problem. Just prohibit banks from getting bigger that the maximum limit (which would obviously be indexed). Banks in violation of the size constraint would be obliged to spin some pieces off in a timely manner. Like a constraint on the size of munitions factories, this would directly limit the risks to society at the expense of lost scale economies. In banking, however, a large empirical literature on economies of super-scale has produced mixed results. Thus, it is not obvious that a size limit of this nature would result in significant efficiency losses.
We have seriously considered an alternative approach; to identify the TBTF banking firms according to specific operating characteristics, and then regulate them much more stringently than other banks. That would be less draconian than a size limit to be sure, but invites gaming the system. Moreover, the regulations on TBTF banks would necessarily be severe—since our intention to render them permanently fail-safe.
Unfortunately, the two reforms we propose have an obvious and potentially fatal flaw. They are mostly backward looking, and address problems that led to the current crisis. The next crisis will surely be different. Moreover, our proposals primarily impose constraints instead of changing incentives. Banks would still have a powerful incentive to become TBTF if they could figure out how to do it. If they became TBTF, the markets would know that and pass along the substantial funding cost advantages that go with de facto government guarantees. There is one possible solution that we immediately dismiss; the government commits to never rescuing failing banks. That is arguably not good policy (why have a central bank?) and an announced policy of this kind would be blatantly time-inconsistent.
We can think of a number of ways in which bankers might game the proposed new system. One way would be to create new updated versions of the chain arrangements—e.g. mortgage bankers, securitization, financial engineering, investment banks, etc.—that led to the impotency of bank regulation over the last decade. Another possibility would be to create very large banking organizations composed of many small, essentially-identical parts. We are thinking of a franchising kind of arrangement. Undoubtedly, there are other possibilities and market participants would have the incentive to discover and exploit them.
The clear implication is that the reforms we propose cannot be static. They must create a framework for dynamic adjustment to evolving financial market reality. We propose the creation of a systemic risk oversight board to monitor such changes and implement mid-stream corrections. Of necessity, this would be a powerful body and its decisions could be controversial and politically contentious. For those reasons, the board should be composed of not just the new regulator, but also representatives from the Administration and the Congress. This broad representation would make it harder for special interests to wield undue influence. Also and importantly, it would virtually guarantee that important deliberations were made public.
If all these changes were put in place, for the first time there would be a significant cost to banks that were TBTF. At present, there are only benefits and the large costs of crises are borne by the economy at large. We recognize that our proposals involve a massive change in regulatory procedures, expand the scope of financial regulation, and re-define the playing field for regulators. Given recent events, however, we think that such changes are called for and the attendant costs are worth bearing. Most current proposals, including those emanating from Washington, leave the TBTF problem untouched or worsen it by encouraging financial consolidation. It is time to recognize that big banks, like big munitions factories, can produce big negative externalities. Further, we believe it is neither too late, nor too early, to do something about this problem.